Molecular Control of Oligodendrocyte Development

Abstract

Myelin is a multilayer lipid membrane structure that wraps and insulates axons, allowing for the efficient propagation of action potentials. During developmental myelination of the CNS, oligodendrocyte progenitor cells proliferate and migrate to their final destination, where they terminally differentiate into mature oligodendrocytes and myelinate axons. Lineage progression and terminal differentiation of oligodendrocyte lineage cells are under tight transcriptional and post-transcriptional control. The characterization of a number of recently identified regulatory factors that govern these processes, which are the focus of this review, has greatly increased our understanding of oligodendrocyte development and function. These insights are critical to facilitate efforts to enhance oligodendrocyte progenitor cell differentiation in neurological disorders that disrupt CNS myelin.

Transcriptional and post-transcriptional control of oligodendrocyte formation

During development, oligodendrocyte progenitor cells (OPCs) terminally differentiate and mature into myelinating oligodendrocytes. This differentiation is a tightly controlled process in which extracellular and intrinsic signals regulate transcriptional and physiological changes in the oligodendrocyte lineage cells. Over the past several decades, considerable insight has been gained into the molecular control of oligodendrocyte development, including the identification of myriad molecular cues that control these processes. These include extrinsic extracellular signals [1] as well as oligodendrocyte-intrinsic transcription factors [2], epigenetic modulators [3–5], microRNAs (miRNAs) [6], and signaling pathways [7].

In addition to the critical role that oligodendrocyte differentiation plays during development, newly formed myelinating oligodendrocytes are also generated in adults (see Box 1). Importantly, demyelinated axons are remyelinated by oligodendrocytes derived from adult OPCs, and remyelination failure results from a diminished capacity to produce myelinating oligodendrocytes [8]. In demyelinating diseases such as multiple sclerosis (MS), insufficient remyelination correlates with disease progression [9]. A full understanding of the factors that control oligodendrocyte differentiation during development and in adults is therefore of utmost importance. Here we review the most recent findings regarding the molecular control of oligodendrocyte development at the transcriptional, post-transcriptional, translational and post-translational levels.

Box 1. Non-developmental myelination in adults.

Myelination is a developmental process that begins around birth in rodents and is mainly completed by early adulthood. Therefore, myelination has been long thought to be a primarily developmental process, and myelin considered a nearly static structure in adults. Nevertheless, newly formed oligodendrocytes in the adult brain, derived from adult OPCs, can remyelinate demyelinated axons [9]. In addition, it has recently become increasingly evident that adult-born oligodendrocytes and myelin are crucial for motor-skill learning [94,95], and that myelin is a fairly dynamic structure that undergoes changes and remolding in adulthood [96]. These changes occur at least in part in response to neuronal activity, and are observed in behavioral settings such as skilled learning and social isolation [97]. The unique molecular cues that govern non-developmental CNS myelination, however, remain largely unknown.

Transcriptional control of oligodendrocyte development

Transcriptional control of oligodendrocyte development by transcription factors

During development, oligodendrocyte lineage cells arise from neuroepithelial cells in the ventricular zone. Once the lineage of each oligodendrocyte is specified, tightly-controlled feedback loops involving several transcription factors work to either maintain proliferating OPCs in their progenitor state and prevent their premature differentiation, or release them for differentiation into mature oligodendrocytes (reviewed thoroughly in [2]).

Newly arisen oligodendrocyte lineage cells express the transcription factors NK6 homeobox 1 (NKX6.1), NK6 homeobox 2 (NKX6.2), NK2 Homeobox 2 (NKX2.2) and Oligodendrocyte transcription factor 2 (OLIG2). OLIG2 and NKX2.2 are important determinants of oligodendrocyte differentiation and must be co-expressed as a precondition for differentiation to occur. In the developing neural tube, however, they are expressed in adjacent domains of the ventral ventricular zone and repress one another [10,11].

OLIG2 directly induces the expression of SRY-box 10 (SOX10), a transcription factor that is critical for oligodendrocyte maturation [12] (see Figure 1, Key Figure). In turn, the positive feedback loop between SOX10 and OLIG2 maintains OLIG2 expression in SOX10-expressing cells [13]. The calcineurin-mediated activation of the nuclear factor of activated T cells (NFAT) protein NFATC2 represses the inhibitory effect of OLIG2 on SOX10-mediated NKX2.2 expression and vice versa; NFATC2 also relieves the inhibitory effect of NKX2.2 on SOX10-mediated OLIG2 expression. Hence, calcineurin-mediated activation of NFATC2 allows for co-expression of OLIG2 and NKX2.2, which in turn results in the initiation of oligodendrocyte differentiation [14].

Figure 1, Key Figure: Main axes of SOX10-mediated oligodendrocyte differentiation.

Four main SOX10 axes have been recently characterized: the CHD8-BRG1-OLIG2 axis, the Calcineurin-NFAT axis, the MYRF axis and the TCF7L2 axis. The proteins that are involved in this circuit are marked in cyan, myelin transcripts are marked in yellow, and post-translational modifications (de-phosphorylation of NFAT and ZFP24 and formation of MYRF-homotrimer) are marked in magenta.

The role of SOX10-interacting proteins in oligodendrocyte differentiation

Full differentiation will not occur, however, without the interaction of SOX10 with a number of other proteins, such as myelin regulatory factor (MYRF) [15,16]. Only very recently has SOX10 been recognized as a major determinant in the terminal differentiation of oligodendrocytes. Four main axes of SOX10-mediated oligodendrocyte differentiation have been characterized: the MYRF axis, the chromodomain helicase DNA binding protein 8 (CHD8)-SWI/SNF related, matrix associated, actin dependent regulator of chromatin, subfamily A, member 4 (BRG1)-OLIG2 axis, the transcription factor 7 like 2 (TCF7L2) axis, and the Calcineurin/NFAT axis (see figure 1).

SOX10 was established early on as a required protein for the maturation of oligodendrocytes and as a direct activator of several genes critical to the myelination process [17], including the transcription factor Myrf [16]. Once induced, MYRF mediates the progression of pre-myelinating oligodendrocytes to a mature, myelinating state [18]. Both MYRF and SOX10 bind many of the same transcriptional control regions in proximity to genes critical to the myelination process, though studies have shown that they are also able to bind individual enhancers independently [16,19]. SOX10 binds to the active enhancer of the pro-myelinating transcription factor zinc finger protein 24 (Zfp24, formerly known as Zfp191) [20]. In turn ZFP24 binds to the enhancer regions of Sox10 and Myrf and increases their expression [21].

The chromatin remodeler CHD8 activates the expression of BRG1-associated SWI/SNF complexes [22]. BRG1, together with OLIG2, activates the expression of ZFP24 [23] and initiates the expression of the downstream chromatin remodeler chromodomain helicase DNA binding protein 7 (CHD7) [24]. In turn, CHD7 cooperates with SOX10 and regulates the onset of CNS myelination and remyelination [25].

At the onset of differentiation, TCF7L2 interacts with transcriptional co-repressor zinc finger and BTB domain-containing protein 33 (ZBTB33) to block β-catenin signaling, which impedes myelination [26]. During oligodendrocyte lineage progression, TCF7L2 cooperates with SOX10 to induce the expression of other genes crucial for oligodendrocyte differentiation, such as Zfp24, Oligodendrocyte transcription factor 1 (Olig1) and Myrf. In addition, TCF7L2 directly activates genes important for myelination and cholesterol biosynthesis [27].

Despite the ongoing discovery of numerous factors and extracellular signals that control oligodendrocyte differentiation, the full integration of this data into a clear understanding of the extracellular factors that are upstream of SOX10 activation in the CNS remains incomplete. In the PNS, Neuregulin on the surface of the axon signals through the ErbB receptor tyrosine kinases on Schwann cells to control myelination [28]. Neuregulin-ErbB signaling results in an increase in intracellular calcium levels, which activates Calcineurin, a calcium- and calmodulin-dependent phosphatase, to mediate the dephosphorylation of NFATC4. Once dephosphorylated, NFATC4 cooperates with SOX10 to drive the expression of the pro-myelination factor early growth response 2 (KROX20) which subsequently leads to PNS myelination [29]. A significant step in our understanding of CNS myelination was recently achieved by the discovery that calcineurin-dependent dephosphorylation of NFATC2 results in its activation [14]. In turn, activated NFATC2 cooperates with SOX10 to release the repression of OLIG2 on NKX2.2, and vice versa [14]. Nevertheless, the triggering event that leads to the intracellular calcium rise that activates calcineurin in oligodendrocyte lineage cells still needs to be determined.

As described above, the major determinants of oligodendrocyte differentiation and myelination are the transcription factors OLIG2, SOX10, NKX2.2, ZFP24 and MYRF. Together they comprise the core regulatory network that controls oligodendrocyte differentiation. This network is summarized in figure 1. Nonetheless, many other transcription factors that play a role in oligodendrocyte lineage cells have been identified over the years; their role in oligodendrocyte lineage cells, myelination and remyelination is summarized in Table 1.

Table 1:

Transcription factors that play a role in oligodendrocyte development and a brief summary of their role in oligodendrocyte lineage cells, myelination, and remyelination.

Protein	Role in OL development	Effect of its ablation	Role in remyelination
ASCL	Plays a role in OPC formation [104] and OL differentiation [105].	Reduced number of OPCs [104] and reduced number of OLs [105]	Required for proper remyelination [106]
HES1	Maintains cells as OPCs and inhibits differentiation [107].	ND	ND
HES5	Maintains cells as OPCs and inhibits differentiation [107].	Premature OPC differentiation [108]	ND
ID2	Maintains cells as OPCs and inhibits differentiation [109]	Premature OPC differentiation [110]	ND
ID4	Maintains cells as OPCs and inhibits differentiation [109]	Premature OPC differentiation [111]	ND
KLF9	Promotes OPC differentiation in vitro [112].	No effect [112].	Important for remyelination [112].
MYRF	Crucial for OPC differentiation [18]	Normal OPCs and loss of OLGs [18].	Important for remyelination [113]
MYT1	Plays a role in OPC proliferation and OL differentiation [114]	ND	ND
NFAT proteins	Promotes OPC differentiation [14]	OL lineage specific ablation of Calcineurin inhibits OPC differentiation [14]	ND
NFIA	Controls OPC formation [115]	ND	Inhibits OPC differentiation during remyelination [116]
NKX2.2	Plays a role in OPC differentiation [117].	No effect on OPCs, loss of OLs [117].	ND
NKX6.1	Play a role in OPC formation [118]	Reduced number of OL lineage cells [119]	ND
NKX6.2	Plays a role in OPC formation [118]	Paranodal defects [120]	ND
OLIG1	Delayed OPC differentiation [121]	Depends on the experimental system: Normal numbers of OPCs and reduced number of OLs [122] or no change in OL numbers [121]	Important for remyelination [123]
OLIG2	Important for OL lineage cells [124]	Ablation at the OPC stage results in hypomyelination, while ablation at premature OL stage expedites myelination [125]	Olig2 gain of function at the OPC stage promotes remyelination [126]
SOX2 and SOX3	Play a role in OL differentiation [127]	Impairs OL differentiation [127]	ND
SOX4	Promote OPC differentiation [128]	Lethal, however no overt CNS defects in the embryo [129]	ND
SOX5 and SOX6	Inhibit OPC differentiation [130]	Premature OPC differentiation [130]	ND
SOX8	Controls gliogenesis [131]	Loss of OPCs [131].	ND
SOX9	Controls gliogenesis [132].	Loss of OPCs [132].	ND
SOX10	Plays a role in OL differentiation [133]	Normal OPCs, loss of OLs [133]	ND
SOX17	Plays a role in OPC differentiation [134,135]	ND	Overexpression increases the number of OLs and improves remyelination [136]
SREBP proteins	Promote OPC differentiation [137]	No change in OPCs, educed number of OLs [36]	ND
TCF7L2	Promotes OPC differentiation [27,43,138]	Reduced number of OLs [27,43,138]	ND
YY1	Plays a role in OL differentiation [139].	Prevents OL differentiation [139].	ND
ZBTB7A	Modulates the expression of transcripts important for myelination [140].	Normal numbers of OPCs and OLs [140].	Important for remyelination [140].
ZEB2	Plays a role in OPC differentiation [141].	Reduced number of OLs [141].	ND
ZFP24	Plays a role in OL differentiation [21]	Hypomyelination with normal numbers of OPCs and OLs [77]	ND
ZFP488	Promotes OPC differentiation [142]	ND	Promotes OPC differentiation during remyelination [143]

Legend: OL, oligodendrocyte; ND, not determined.

A class of transcription factors that has been gaining attention in this context is nuclear receptors. These are ligand-activated transcription factors that work through homo dimerization or dimerization with each other and by interacting with co-repressors or co-activators [30]. Nuclear receptors including the retinoid X receptors (RXR), vitamin D receptors, liver X receptors (LXR), Thyroid hormone receptors (THR), peroxisome proliferator-activated receptors (PPAR), and the chicken ovalbumin upstream promoter-transcription factors (COUP-TF), are known to play a role in oligodendrocyte development and myelination (see table 2). In addition to their role in oligodendrocyte lineage cell development RXR and LXR play a fundamental role in myelin debris clearance following demyelination, which is a limiting factor in remyelination of aged animals [31,32].

Table 2:

Nuclear receptors that play a role in oligodendrocyte development and a brief summary of their ligands and their roles in oligodendrocyte lineage cells, myelination, and remyelination.

Nuclear Receptor Family	Ligand	Family member name	Effect of its ablation	Role in OL development	Role in remyelination
Chicken ovalbumin upstream promoter-transcription factor	Orphan receptor	COUP-TFI	Hypomyelination [144]	Important for OPC differentiation [144]	ND
Liver X receptors	Oxysterols	LXRα	Accumulation of myelin debris following demyelination [32] and hypomyelination [145]	ND	Important for remyelination [32,145]
		LXRβ	Hypomyelination [145,146]	Important for formation of OL lineage cells [145,146]	Important for remyelination [145,146]
Peroxisome proliferator-activated receptors	15-Deoxy-delta-12,14-prostaglandin J2	PPARγ	ND	Important for OPC differentiation [147]	Important for remyelination [148]
		PPARδ	ND	Important for OPC differentiation [149]	ND
Retinoid X receptors	9-cis-retinoic acid	RXRα	Accumulation of myelin debris following demyelination [31]	ND	Important for remyelination [31]
		RXRγ	Impaired OPC differentiation [150]	ND	Important for remyelination [150]
Thyroid hormone receptors	T3	THRα and THRβ	Myelin defects and sustained elevated numbers of OPCs in adults [151]	Important for OPC differentiation [151]	ND
Vitamin D receptor	Vitamin D	VDR	ND	Induces OPC differentiation [152]	Important for remyelination [152]

Legend: OL, oligodendrocyte; ND, not determined.

Signaling pathways and metabolic control of oligodendrocyte development

Myelin is a lipid-rich structure. It is therefore not surprising that myelination relies heavily on lipid synthesis [33], and that hypomyelination is characterized by reduced expression of transcripts important for lipid synthesis [34]. Interestingly, lipids may play a greater role in oligodendrocyte development beyond simply serving as the building blocks of the multilayered myelin sheath. High-throughput pharmacological screens have identified myriad small molecules that have the capacity to enhance oligodendrocyte maturation and CNS myelination. Unexpectedly, it was recently discovered that a number of these “hits” function by targeting the cholesterol biosynthesis pathway, resulting in the accumulation of 8,9-unsaturated sterols in oligodendrocytes [35]. The mechanisms by which 8,9-unsaturated sterols exert their function, however, remain unknown.

The expression of genes involved in fatty acid and cholesterol metabolism is controlled by the transcription factors sterol regulatory element binding proteins (SREBPs), which are encoded by the genes Srebf1 and Srebf2. Recent studies suggest that these transcription factors have a cell autonomous role in both oligodendrocytes, where they function in myelin lipid synthesis, and in astrocytes, to supply myelin building blocks for the oligodendrocytes through transcellular transport [36]. Lipid synthesis and the expression of SREBPs is controlled by the mammalian target of rapamycin (mTOR) pathway, which plays a fundamental role in oligodendrocyte development, myelination, and remyelination (for a detailed review of the function of mTOR in myelination, see [37]).

The Raf-MAPK-ERK1/2 pathway is also known to control oligodendrocyte differentiation and myelination. Ablation of both extracellular signal-regulated kinase 1 and 2 (ERK1 and ERK2) results in hypomyelination, but characterization of the upstream mediators of this pathway was until recently lacking [38,39]. A recent finding has helped clarify this process: the G protein-coupled receptor GPR37 was shown to be a mediator of the Raf-MAPK-ERK1/2 pathway in oligodendrocytes [40]. By tightly controlling ERK phosphorylation and nuclear translocation in oligodendrocyte lineage cells, GPR37 was found to inhibit the premature differentiation of oligodendrocytes.

The recently identified factors that govern the transcriptional control of oligodendrocyte differentiation expand our knowledge of the tight transcriptional control governing lineage progression and gene expression in oligodendrocyte lineage cells. This data is crucial for our attempts to intervene and enhance the myelination and remyelination processes in the CNS (see Box 2).

Box 2. Pushing oligodendrocyte progenitor cells toward terminal differentiation as a therapeutic possibility.

The CNS has a robust potential to remyelinate demyelinated axons, as exemplified in diseases such as MS. During the remyelination process, OPCs in the vicinity of the demyelinated areas migrate to the lesion site, differentiate into mature oligodendrocytes, and remyelinate the demyelinated axons. With increasing age, however, the capacity for myelin repair is diminished, resulting in severe progression of the demyelinating disease [9]. The currently available MS therapies focus on modulation of the immune system. Nevertheless, the brains of MS patients with remyelination failure are characterized by the presence of oligodendrocytes at the pre-myelinating stage in the vicinity of demyelinated axons [8]. These oligodendrocytes send processes to the axons, but fail to fully differentiate and myelinate the axons. Large pharmacological screens have therefore been used to find small molecules that can “push” oligodendrocyte lineage cells toward terminal differentiation in demyelinating diseases. Surprisingly, as mentioned above, many of the identified drugs share a common mechanism in which they inhibit enzymes in the cholesterol biosynthesis pathway, resulting in accumulation of 8,9-unsaturated sterol in the oligodendrocyte lineage cells [35]. The underlying mechanism by which the accumulation of this lipid leads to enhance oligodendrocyte maturation remains unknown. The drug clemastine fumarate that enhances OPC maturation is, so far, the only drug that has been shown to be effective in inducing remyelination in human patients [98].

In addition to MS, other neurological conditions would likely benefit from therapeutic approaches that enhance oligodendrocyte maturation. For example, hypoxic brain injury results in focal damage to white matter tracts. OPCs are recruited to areas of injury but fail to fully remyelinate the demyelinated lesions [99]. Therefore, driving oligodendrocyte maturation would likely be therapeutic [100,101]. Pushing oligodendrocyte progenitor cells toward terminal differentiation may also be beneficial following neuronal injury, where remyelination of regenerated axons appears to be a limiting factor [102].

Epigenetic regulation of oligodendrocyte development

Epigenetic regulation drives oligodendrocyte lineage progression and myelination [4]. The chromatin of OPCs, which are dividing cells, is accessible for transcription factors, co-activators or co- repressors and histone-modifying enzymes. At the OPC stage, extracellular cues drive mainly inhibitory pathways that prevent differentiation and promote proliferation. At the onset of the differentiation process, histone deacetylase activity is associated with formation of the heterochromatin and down regulation of the inhibitors that prevent differentiation [41–43]. Epigenetic regulation of oligodendrocyte maturation by DNA and histone modifications and ATP-dependent chromatin remodeling complexes has been recently reviewed thoroughly [22].

Long noncoding RNAs

Long noncoding RNAs (lncRNAs) are a sub-group of noncoding RNAs that are longer than 200 nucleotides but do not translate to proteins. Over 800 lncRNAs are expressed in the CNS, some of which are expressed in a cell type-specific manner [44]. Recently, a dynamic lncRNAs expression database of different stages of oligodendrocyte development was generated [45]. Of the oligodendrocyte-specific lncRNAs, the functional expression of lncOL1 was further characterized. lncOL1 interacts with a component of polycomb repressive complex 2 (SUZ12) to inhibit the expression of oligodendrocyte differentiation inhibitors, thereby promoting oligodendrocyte differentiation [45]. In addition, lnc158 was found to promote oligodendrocyte differentiation by controlling the levels of its target transcript nuclear factor-IB (NFIB) [46]. The functional roles of the remainder of the oligodendrocyte lineage cell-specific or enriched lncRNAs remains to be determined.

MicroRNAs

MicroRNAs (miRNAs) are a sub-group of noncoding RNAs that are approximately 21–25 nucleotides in length. miRNAs are initially transcribed as part of long transcripts and mature through cleavage by the enzyme Dicer. Oligodendrocyte-specific ablation of Dicer revealed that miR-219 and miR-338 are oligodendrocyte-specific miRNAs that promote oligodendrocyte differentiation [47–49]. Recent systematic characterization of miR-219 targets in oligodendrocyte lineage cells revealed that this miRNA has a cell autonomous role in oligodendrocyte lineage cells. MiR-219 has been found to repress known inhibitors of oligodendrocyte differentiation such as Lingo1, in a stage-specific manner [50]. In addition, these studies have revealed mir-219 targets that were previously not known to play a role in oligodendrocyte differentiation, such as the transcription factor ETS variant 5 (ETV5). This transcription factor was shown to promote oligodendrocyte differentiation in vitro, suggesting that miR-219 exerts its transcriptional activity, at least partially, by mediating ETV5 activity. Oligodendrocyte specific ablation of miR-219 resulted in hypomyelination, but to a lesser extent than ablation of Dicer, suggesting that additional miRNAs are important for CNS myelination. Unlike miR-219, ablation of miR-338 in oligodendrocytes did not result in overt effects on oligodendrocyte differentiation or myelination [50]. Nevertheless, it has been shown in vivo that miR-219 and miR-338 work in collaboration to promote oligodendrocyte differentiation and myelination [50]. Recently, miR-125a-3p was shown to prevent premature differentiation of oligodendrocyte progenitor cells [51]. The predicted miR-125 targets SMAD Family Member 4 (SMAD4), Tyrosine-Protein Kinase Fyn (FYN), Ras homolog family member A (RHOA), Mitogen-Activated Protein Kinase 1 (P38) and Neuregulin 1 (NRG1), which are known to play a role in oligodendrocyte development, suggest that miR-125 indirectly affects the expression of MBP [51].

Recently, it has also become clear that miRNAs have an exciting therapeutic potential; specifically, mir-146 was shown to promotes oligodendrocyte differentiation and remyelination in various models of CNS demyelination [52–54]. The mechanism of miR-146 action may be meditated by the known miR-146 target interleukin-1 receptor-associated kinase 1 (IRAK1) [52–54]. Mir-219 was shown to promote remyelination in the lysolecithin and cuprizone models of toxicant-induced demyelination, as well as in the experimental autoimmune encephalomyelitis (EAE) model of inflammatory demyelination [55,56]. In addition, serum exosomes containing miR-219 and dendritic cell exosomes enriched in miR-219, miR-9 and miR-92–1, which are known to play a role in CNS myelination (see table 3), were shown to improve myelination and remyelination [57,58], suggesting that exosome-mediated delivery of micro-RNAs has therapeutic potential [59,60].

Table 3:

List of miRNAs that play a role in oligodendrocyte development and a brief summary of their targets, and their roles in oligodendrocyte lineage cells, myelination, and remyelination.

Name	Role in OL lineage cells	Effect of its ablation	Targets/mechanism of action	Role in remyelination
miR-219	Important for OPC differentiation [47–50]	OL specific ablation results in hypomyelination [50]	Inhibits the expression of inhibitors of OPC differentiation, such as Etv5, Lingo-1, NFIA and PDGFRα [50]	Promotes remyelination [50,55,56]
miR-338	Important for OPC differentiation [47–49]	OL specific ablation did not result in overt phenotype [50]	ND	ND
miR-138	Important for OPC differentiation [47–49]	ND	ND	ND
miR-23a	Important for OPC differentiation [153,154]	Hypermyelination in miR-23a overexpressing mice [154]	By controlling the expression of its targets lamin B1, phosphatase and tensin homolog on chromosome 10 (PTEN) and the lncRNA 2700046G09Rik miR-23a controls OPC differentiation. [154]	ND
miR-17–92 cluster; miR-17, miR18a, miR-19a, miR-19b-1, miR-20a, and miR-92–1	Regulates OPC proliferation [155]	Hypomyelination [155]	Controls OPC proliferation by regulating Akt signaling, mediated by the predicted target PTEN [155]	ND
miR-297c-5p	Important for OPC differentiation [156]	ND	By targeting cyclin T2 (CCNT2) induces OPC differentiation [156]	Promote remyelination [156]
miR-146	Important for OPC differentiation [52,54]	ND	By repressing its target IL-1 receptor-associated kinase 1 (IRAK1) promotes OPCs differentiation. [52,54]	Promotes remyelination [52,53]
miR-9	Important for OPC differentiation [157]	ND	Prevents expression of PMP22 protein in OPCs [157].	ND
miR-125	Maintains cells as OPCs [51]	ND	The predicted targets Smad4, FYN, RHOA, P38 and NRG1 suggest that mir-125 may indirectly affect MBP expression [51]	ND

Legend: OL, oligodendrocyte; ND, not determined

A number of other miRNAs that play a role in oligodendrocyte lineage cells have been identified over the recent years; their role in oligodendrocyte lineage cells, myelination, and remyelination is summarized in Table 3.

Translational control of oligodendrocyte development

Global protein translation – the role of eukaryotic translation initiation factors

During developmental myelination or during remyelination, oligodendrocytes produce vast amounts of membrane proteins and lipids that form the myelin sheath through the secretory pathway. The production of such vast amounts of protein makes oligodendrocytes uniquely vulnerable to changes in protein translation homeostasis [61]. During developmental myelination, the eukaryotic translation initiation factor 2B (eIF2B) has an oligodendrocyte cell-autonomous role in maintaining protein translation homeostasis [62]. Mutations in eIF2B inhibit protein translation, delay recovery during the integrated stress response, and cause Vanishing White Matter Disease, an inherited autosomal-recessive CNS hypomyelinating disease [63]. Inhibition of global protein translation mediated by the eukaryotic translation initiation factor 2 alpha (eIF2alpha) phosphorylation protects oligodendrocytes from an inflammatory environment, increases their survival, and enhances myelination [64,65], indicating that oligodendrocytes are specifically sensitive to proper homeostasis of protein translation. Inhibition of global protein translation in order to protect oligodendrocyte lineage cells during remyelination is a potential new direction for treating demyelinating diseases (see Box 3).

Box 3. Protecting oligodendrocytes from CNS inflammation by an enhanced integrated stress response as a therapeutic horizon.

During developmental myelination or during remeylination, oligodendrocytes turn into “protein factory cells” that synthesize vast amounts of protein at an estimated rate of about 10⁵ proteins per minute [103]. This makes oligodendrocytes uniquely vulnerable to changes in protein translation homeostasis. Numerous studies have shown that the inflammatory condition in the brain during demyelinating diseases activates ER stress in oligodendrocytes, which is deleterious for the cells (for review, see [61]). Preclinical work in murine models of demyelination has shown that inhibition of global protein translation and enhancement of the integrated stress response, by either genetic manipulations or pharmacological means, can protect oligodendrocytes and improve their myelination function [64,65].

Spatial and temporal control of protein translation in oligodendrocytes.

One of the unique features of oligodendrocytes is that axonal myelination depends heavily on the uneven distribution of myelin proteins within the oligodendrocyte. Specifically, myelin basic protein (MBP) mediates membrane compaction [66]. Therefore, MBP translation, which may damage cell membranes at the soma, requires anterograde mRNA transport using the dynein/dynactin motor complex [67] followed by local translation of Mbp mRNA in the cell’s processes, in proximity to the developing myelin sheath [68]. The factors that participate in the spatial and temporal control of MBP translation are largely unknown. The ERK/MAP kinase effector ERK2 has been shown to control the translation, but not the transcription, of MBP [69]. Similarly, the tumor overexpressed gene (TOG) was shown to control the translation, but not the transcription, of MBP, most likely by affecting its mRNA translocation to the cell processes [70]. Several mRNA-binding proteins from the heterogeneous nuclear ribonucleoprotein (hnRNP) protein family have been shown to have a specific role in local translation of MBP. For example, hnRNP-A2 is responsible for MBP mRNA transport, while hnRNP-E1 inhibits its translation and hnRNP-K stimulates its translation [71]. HnRNP-A2 exert its activity by binding to N⁶-methyladenosine (m⁶A)-bearing RNAs [72], suggesting that mRNA methylation may play a role in oligodendrocyte development. Nevertheless, the role that the m⁶A mark plays in oligodendrocyte development and CNS myelination remains unknown (see Outstanding Questions). Surprisingly, transcriptome studies on biochemically-isolated myelin from mouse brains have revealed that myelin is enriched in a large number of mRNA molecules, including transcripts that are important for myelin sheath formation as well as those important for local translation, suggesting that this spatial and temporal control of protein translation may be relevant to many other myelin transcripts [73].

In vivo, the spatial and temporal control of CNS myelination depends on programmed cell death in unmyelinated brain regions [74,75], and on the ability of the surviving cells in myelinated brain regions to properly interact with axons [76]. The lysosomal G protein ras related GTP binding A (RAGA) promotes myelination by inhibiting the expression of the transcription factor EB (TFEB), which acts as a myelination inhibitor [75]. TFEB controls the expression of ER stress and pro-apoptotic genes in premyelinating oligodendrocytes, in a cell autonomous manner. These TFEB induced genes mediate oligodendrocyte programmed cell death in unmyelinated brain regions, thereby shaping CNS myelination [74,75]. Furthermore, proper myelin targeting to axons (and not to the cell soma), membrane wrapping, and internode extension are heavily dependent on proper axon-glia interaction mediated by the neuronal cell adhesion molecules 2 and 3 (CADM2 and3) and on the oligodendrocyte cell adhesion molecule 4 (CADM4) [76].

Post-translational control of oligodendrocyte development

Post-translation modification of proteins, such as phosphorylation, adds another layer of complexity to the molecular control of oligodendrocyte development. De-phosphorylation of NFAT proteins, mediated by Calcineurin, is the key event that initiates both PNS and CNS myelination [14,29].

Dephosphorylation of the transcription factor ZFP24 in its DNA binding domains increases ZFP24 binding to its DNA targets in proximity to genes that are crucial for oligodendrocyte differentiation and myelination [21]. The binding of ZFP24 to its DNA binding targets is important for the expression of numerous genes that are critical for CNS myelination [34] and is a key event in CNS myelination, given that Zfp24 null mice are severely hypomyelinated [77]. Human patients hemizygous for 18q chromosomal deletions that encompass Znf24 display seizures and tremors, suggestive of myelin abnormalities [78]. The enhancement of ZFP24 activity by modulation of its phosphorylation state may induce oligodendrocyte differentiation and therefore may be beneficial to patients suffering from demyelinating diseases (see Box 2). Nonetheless, the kinases and phosphatases that regulate ZFP24 activity remain unknown, as are the upstream modulators of this control.

OLIG1 and OLIG2 also undergo phosphorylation, and the phosphorylation state of OLIG2 has been shown to determine its function. Dephosphorylating serine 147 (S147) of OLIG2, for example, dictates the formation of oligodendrocyte lineage cells [79]. Meanwhile, phosphorylation of OLIG2 on residues S10, S13 and S14 is elevated during the neural progenitor stage and is reduced as the cells differentiate to myelinating oligodendrocytes. Likewise, the kinase glycogen synthase kinase 3 (GSK3) mediates OLIG2 phosphorylation at S10, cyclin-dependent kinases 1 and 2 (CDK1 and CDK2) at S14, and casein kinase 2 (CK2) at S13, to determine OLIG2 functions [79–81]. The phosphatase(s) that mediate OLIG2 dephosphorylation are as yet unknown.

OLIG1 is phosphorylated in oligodendrocyte lineage cells at S149 by protein kinase A (PKA) [82]. Independent of its phosphorylation status, OLIG1 is also acetylated at lysine 150 in vivo [83]. OLIG1 acetylation is regulated by the acetyltransferase CREB-binding protein (CBP), and OLIG1 deacetylation is regulated by histone deacetylases HDAC1, HDAC3, and HDAC10. Acetylation of OLIG1 regulates its chromatin dissociation and subsequent translocation to the cytoplasm, while also increasing its binding to the oligodendrocyte differentiation inhibitor of DNA binding 2 (ID2). The retention of the OLIG1-ID2 complex in the cytoplasm likely relieves the inhibitory effect of ID2 on oligodendrocyte differentiation [83].

MYRF is a membrane-bound transcription factor that is associated with the endoplasmic reticulum (ER). Its activity is controlled by complex post-translational processing – MYRF activation requires self-cleavage, which results in the release of its N-terminal fragment [19]. After cleavage, the N-terminal undergoes homo-trimerization, resulting in the active form of the protein [84]. As a homotrimer, MYRF is then translocated to the nucleus where it acts as a transcription factor that mediates the expression of genes crucial for oligodendrocyte differentiation and myelination [18]. The cleavage of MYRF is controlled by the ER resident transmembrane protein 98 (TMEM98), which binds to the C-terminal of MYRF to prevent its self-cleavage, thereby providing another layer of negative regulation to prevent premature differentiation of oligodendrocyte lineage cells into oligodendrocytes [85].

Concluding Remarks and Future Perspectives

Over the past few decades, numerous factors have been identified that control oligodendrocyte development. These include extrinsic extracellular cues as well as oligodendrocyte-intrinsic transcription factors, epigenetic modulators, DNA methylation, non-coding RNAs and signaling pathways. Many of these factors were originally discovered through large random screens and their functional interactions with each other were poorly understood. Nevertheless, over the last few years, many new components of the regulatory network that controls the development of oligodendrocytes have been uncovered. These findings have revealed integrated, complex transcriptional networks that control oligodendrocyte differentiation and CNS myelination. Among other advances, these findings have led to a new understanding of the factors that cooperate with SOX10 to mediate CNS myelination, including TCF7L2 [27], CHD7 [25], ZFP24 [21], OLIG2 [16], MYRF [86], and NFATC2 [14], as well as new insights into the role of the calcium- and Calcineurin-mediated network in the initiation of both PNS and CNS myelination [14,29] (see figure 1). This latter finding is of particular interest as recent studies have shown that neuronal activity raises intracellular calcium levels in oligodendrocyte lineage cells and that high frequency calcium transients mediate myelin sheath elongation [87,88]. The concept of calcium-dependent intrinsic transcriptional control of oligodendrocyte differentiation may provide a framework to understand the transcriptional control of activity-dependent myelination. Further studies are required to determine whether the newly identified Calcineurin- NFATC2-SOX10 axis activation in oligodendrocytes is downstream of neuronal activity, and whether there is a direct link between neural activity to the intrinsic transcriptional control of oligodendrocyte differentiation.

The work reviewed here demonstrates that highly complex regulatory networks interact to control oligodendrocyte development and CNS myelination. We expect that this information will serve as the basis for the development of a “tool kit” to treat demyelinating diseases. New therapeutic approaches will likely include strategies to protect pre-existing oligodendrocytes from stress (see Box 3), improve myelin clearance [31,32], and enhance oligodendrocyte differentiation (see Box 2). We anticipate that these strategies will improve remyelination and the overall outcome of hypomyelinating and demyelinating diseases of the CNS.

Despite our increasingly detailed understanding of oligodendrocyte maturation in the context of developmental myelination, our understanding of oligodendrocyte development in adulthood and disease is still lacking. Specifically, in the context of disease, oligodendrocyte lineage cells can give rise to Schwann cells [89,90] or become antigen presenting cells [91]. In addition, post-mitotic oligodendrocytes have recently been shown to participate in CNS remyelination [92]. Moreover, for unknown reasons, adult-born myelin sheaths are shorter and thinner than myelin sheaths that are deposited during development [93]. The transcriptional and post-transcriptional controls of these phenomena remain poorly understood, and represent one of many fascinating areas for future work (see Outstanding Questions).